Answers are grounded in this paper and cite their sources.
Joint source-channel coding (JSCC) for semantic communication (SemCom) has achieved significant progress. However, due to the degradation of underwater images, directly using JSCC for underwater SemCom leads to inadequate semantic extraction. To this end, this article proposes a transmission map-guided JSCC (TGJSCC) for underwater SemCom to better extract and transmit the semantic information of underwater degradation images, called TGJSCC. Specifically, we design the TGJSCC encoder to extract abundant semantic information of underwater degraded images. TGJSCC encoder first uses the transmission map generated by the underwater imaging model to help JSCC locate the focal regions in underwater degraded images, and then computes the global information in the latent space to obtain abundant semantic information. To transmit semantic information over the limited underwater channel, the semantic importance compression module (SICM) is proposed to compress semantic information while retaining useful information. Finally, the TGJSCC decoder is designed to reconstruct raw underwater degraded images from the semantic information transmitted by the underwater channel. The experimental results and analysis demonstrate that compared with the traditional separation source-channel coding (SSCC) methods and JSCC methods, the underwater SemCom based on TGJSCC not only extracts abundant semantic information of underwater degradation images, but also recovers the high-precision images.
No takes yet. Share an insight, caveat, or question.
Cheng et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: